Low-budget satellite tracking system for highly elliptical orbits

12/17/09; Thomas.Rode@gmx.de

The background:

Back in 2003 I was a member of the club station DL0AMD. In those days AO-40 was
still active. So the idea came up to upgrade DL0AMD by adding equipment for
satellite communication. Donators were found for almost everything (transceiver,
Yagi, dish, …) but the tracking system, namely the elevation/azimuth rotor. The only
hardware available was a simple, one axis rotor. After looking a little deeper into the
“rotor-issue”, I came up with an idea on how to make use of the available hardware
for our special purposes. After installation that system was used for many month (and
QSOs) and it actually worked quite well. Unfortunately our experiment was ended by
the failure of AO-40. But who knows, maybe there will be a successor in the near
future. If that happens and if you are not one of those who can already call a full-size
satellite tracking system their own, you might be interested in just getting started by
taking our approach.

The idea:

In order to aim a TV-satellite dish to its target, a method called “polar mounting” is
often used. That method requires only one degree of freedom (i.e. rotation axis /
drive). The axis is aligned (parallel) to the axis of Earth rather that perpendicular. By
turning that axis, an imaginary circle around the equator (geostationary orbit) will be
described by the beam of the antenna. Mathematically, of course, that circle is only
correctly drawn, if your location is at one of the poles of the planet. But since the
distance between the two axis (Earth / rotor) is small in comparison to the orbital
height (36000km), the error created by it can often be neglected (unless your dish is
the size of a radio telescope).
But how can “polar mounting” be used for satellites in elliptical orbit that do not circle
the equator? Interestingly the principle is not limited to equatorial circles but can be
applied to “scan” any orbit, as long as it is circular and parallel to the equatorial plane
(as for example used for the aiming of astronomical instruments). Such imaginary
circles can also be drawn simply by connecting all apogees of an elliptical orbit, as
they are shifting over time. Knowing in addition that at highly elliptical orbits the
satellite remains near the apogee for most of the time (a phenomenon referred to as
“apogee dwell”), the aiming of the antenna will be accurate enough over a long time
period as well. For a more detailed illustration please see diagram 1.

Axis of rotor
H
or
iz
Meridian plane

on
al
p la
ne
QTH

ε Apo

South

Equatorial plane β
W
λ ι
0°Long.
0°Lat.

Sate
ll ite tr
ack

S Satellite at
apogee

Virtual track of apogee-points

Diagram 1: Principle of polar mounting

Here a sectioned illustration of the Earth is given, showing the equatorial plane as
well as the plane of the prime meridian (longitude 0). The QTH is defined by its
longitude (λ λ) and latitude (β
β ). According to the principle of polar mounting, the rotor is
aligned in parallel to the Earth-axis. When pointed to the south, the antenna(-system)
mounted to it has an elevation (from the plane of horizon) of εApo. Let us assume that
in this position the antenna is pointed to the (exemplary) apogee of a satellite in
elliptical orbit. Since for most orbits the apogee will shift over time (relative to the
observers location), it will create a virtual track of apogees, which is circular, parallel
to the equatorial plane and can be tracked by the (polar mounted) antenna. The
angular distance between that virtual track (plane) and the equatorial plane is called
inclination (ι).
Of course communication is not limited to the time when the satellite is exactly at
apogee. How long the usable time-window actually is, depends on the orbit itself as
well as the beamwidth of the antenna. To give an example, in case of DL0AMD we
used a 16 element x-Yagi for the 70cm uplink (-3dB beamwidth = 30°) and a 60cm
offset-dish for reception at 13cm (-3dB beamwidth = 15°). In practise it turned out
that good communication was possible basically throughout the whole DX-time
window.
Setting up the rotor:

When mounting the rotor it is important that it remains adjustable in both vertical as
well as horizontal direction. In case of DL0AMD we used scaffolding poles and
clamps to insure that necessary flexibility. The next task is now to give the rotor an
orientation that is in parallel to the Earth axis. As a first step, the rotor has to be
turned until the tip of the axis points to the north. Looking quite trivial on first glance,
that part requires some accuracy and is actually a bit tricky. If you have access to a
good map, you could use landmarks to bear the direction. Or you could use the polar
star (of course only if you live in the northern hemisphere). In case of DL0AMD, we
calculated the time for the transit of the sun (when the sun is exactly in the south at
our location) on a certain date. After that time had arrived we took the bearing by
turning the rotor/axis towards the sun. Do not stare into the sun but use some
aids or, as we did, drop a perpendicular from the axis and align to the axis
shadow. Instead of doing the calculation by yourself, you can use some
astronomical software (e.g. the freeware “Home Planet” from www.fourmilab.ch) to
determine when the sun is in transit at your location. The second step is to set the
β) of the QTH. One
necessary vertical tilt of the axis, which is identical to the latitude (β
possible way of doing that adjustment is to use a plumb line. The angle between the
axis and the line then has to be 90°+ β .

Setting up the antenna system:

After the antenna system has been hooked up to the rotor, it should be pointed to the
south. Its elevation (angular distance to the plane of horizon) can now be set to
εApo. The angle εApo has to be calculated from the Kepler elements, using the
parameter „Mean Motion“ (MM), Eccentricity (EX), Arg of perigee (δ δ) as well as
Inclination ι. Additionally the latitude (β
β) of the QTH will be needed again. Kepler
elements are available in two different formats, the so-called NASA-2-line format and
the AMSAT format. For both an exemplary data set from AO-40 is listed below. The
parameters of interest are highlighted.

Set of Kepler elements in AMSAT – format:

Satellite: AO-40
Catalog number: 26609
Epoch time: 03014.42373129
Element set: 304
Inclination: 7.9543 deg Inclination of orbit in Degree
RA of node: 70.5875 deg
Eccentricity: 0.7935115 Eccentricity of orbit
Arg of perigee: 133.0894 deg Argument of perigee in Degree
Mean anomaly: 327.3710 deg
Mean motion: 1.25596681 rev/day Mean motion in revolutions per day
Decay rate: 2.5e-07 rev/day^2
Epoch rev: 1014
Checksum: 282

Same set of Kepler elements in NASA-2-line format:

AO-40
1 26609U 00072B 03014.42373129 .00000025 00000-0 10000-3 0 3042
2 26609 7.9543 70.5875 7935115 133.0894 327.3710 1.25596681 10146
Note that eccentricity in NASA-2-line format is given in positions after decimal point
only. So 7935115 stands for 0.7935115. With sufficient accuracy εApo can be
calculated using equation 1.

3
MM 2 ∗ sin α
ε Apo = 90° − α − arctan
6.636 ∗ (1 + EX ) − cos α
Equation 1

The angle α must be entered in degree and is calculated as follows (equation 2):

α = β + ι ∗ sin (δ ) Equation 2

For DL0AMD the latitude is β = 51.133°. Entering that as well as the exemplary
Kepler data into the two equations, the result is:

α ( DL 0 AMD ) = 51 .133 ° + 7.9543 ° ∗ sin (133 .0894 ° ) = 56 .942 °

ε Apo ( DL0 AMD ) = 90° − 56.942°

1.255966812 ∗ sin(56.942°)
3
− arctan
6.636 ∗ (1 + 0.7935115) − cos(56.942°)

So for DL0AMD we get εApo = 28.15°. Finally it should be mentioned that for
satellites without active orbit control (e.g. by onboard thrusters) the value of εApo will
change over time. Fortunately that change happens very slowly, making corrections
of the antennas elevation a rare event, necessary only if the value has changed by
some degree. At DL0AMD we had a protractor mounted to the antenna system,
making adjustments of εApo a quick and easy job.